{"id":"W4413415478","doi":"10.2196/75279","title":"Leveraging Retrieval-Augmented Large Language Models for Dietary Recommendations With Traditional Chinese Medicine’s Medicine Food Homology: Algorithm Development and Validation","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Traditional Chinese Medicine Studies","field":"Medicine","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Southeast University; National Natural Science Foundation of China","keywords":"Computer science; Precision medicine; Traditional Chinese medicine; Algorithm; Artificial intelligence; Natural language processing; Medicine; Alternative medicine; Pathology","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009685152,0.0003781078,0.0007756658,0.0005357959,0.0003784945,0.00001209304,0.0001561081,0.0001773438,0.0002139738],"category_scores_gemma":[0.0007377623,0.000237029,0.00005856488,0.0006416051,0.0005004021,0.0003084801,0.00007917504,0.0005053655,0.000002987357],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001675775,"about_ca_system_score_gemma":0.0004344779,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000706402,"about_ca_topic_score_gemma":0.00001400112,"domain_scores_codex":[0.9970822,0.00004799062,0.00111243,0.0002583032,0.001092345,0.0004067852],"domain_scores_gemma":[0.9979073,0.0009228442,0.0002369871,0.0002504899,0.0003284033,0.0003539759],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003635103,0.003788823,0.01038564,0.01306466,0.008467561,0.0002634409,0.2947308,0.00008849379,0.0002374607,0.05133693,0.2971975,0.3168036],"study_design_scores_gemma":[0.1474669,0.01571662,0.133677,0.03010507,0.003496592,0.003219261,0.1509441,0.3613516,0.0008092505,0.06711879,0.08281474,0.003280047],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4458178,0.00174556,0.4525783,0.08372122,0.001058938,0.003500962,0.0002771405,0.0004583831,0.01084163],"genre_scores_gemma":[0.8987767,0.0003554974,0.07455359,0.01538946,0.001360615,0.000724112,0.008075885,0.00006260163,0.0007015964],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4529588,"threshold_uncertainty_score":0.9665762,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04443141074045245,"score_gpt":0.3301811419147976,"score_spread":0.2857497311743452,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}